Many important human pathogens including those responsible for malaria, dengue, yellow fever, Zika, West Nile fever, and chikungunya are transmitted by mosquitoes. A functioning ?gene drive? system could fundamentally change our strategies for the control of these disease vectors, by allowing us to drive genetically engineered alleles into vector populations. The recently developed CRISPR/Cas9 gene drive (CGD) system promises a highly adaptable mechanism for this purpose that works by converting heterozygotes for the driver construct into homozygotes in the germ line. However, it remains unclear how well this mechanism would work in wild populations where genetic heterogeneity and mutation will likely give rise to alleles that are resistant to the drive mechanism. Importantly, such resistance alleles will be produced by the drive itself when cleavage is repaired by nonhomologous end joining (NHEJ). The goal of this proposal is to develop CGD constructs with reduced rate of resistance allele formation and to experimentally quantify the factors that are important to the evolution of resistance alleles in large cage populations of the model organism Drosophila melanogaster. In our first aim, we will engineer transgenic Drosophila with several CGD constructs that will allow for rapid assessment of driver, wild-type, and resistance genotypes. Our constructs will disrupt target genes that produce an easily identifiable recessive phenotype. In addition, they will also contain a dsRed gene producing a dominant phenotype, allowing us to distinguish driver alleles from different types of resistance alleles, which may or may not disrupt target gene function. Our CGD constructs will be designed to minimize the rate of resistance allele formation, either through use of multiple gRNAs, use of the male-germline-only ?2-tubulin promoter, or use of a shRNA to silence Lig4 ? an essential component in NHEJ machinery. In our second aim, we will use large cage populations of several thousand flies to study the population dynamics of our CGD constructs over the course of several generations. We will first track the frequency of non-driving disrupted target alleles, which will inform us about their fitness costs and the effective population size in our cages. We will then introduce flies with our CGD constructs at low starting frequency into cages of wild-type flies. Phenotypes will be tracked over several generations and resistance alleles will be characterized by sequencing at the end of the experiment to determine the rate at which resistance arose during the spread of the driver. All experiments will be conducted inside a USDA-inspected arthropod containment facility to prevent escape of transgenic insects. There is good reason for caution in considering the use of gene drive systems for genetic manipulation of wild populations. The parameters derived from our experiments will be critical for modeling the population dynamics of such approaches in the wild, and our large cage system will provide a valuable resource for evaluating future approaches engineered to further suppress resistance.